AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
178 Commits 3 Branches 0 Tags
ccad358bc90bdfe83f56bd7c390306f58f3bc496
Commit Graph

153 Commits

Author SHA1 Message Date
Wenkai Du ccad358bc9 Warm up both out-of-place and in-place collectives (#51) 
[ROCm/rccl-tests commit: 5ee7a08994]
 2023-10-16 12:13:50 -07:00
arvindcheru 2e12f0cfce Update Makefile - HIPCC Path Updated to latest (#45) 
[ROCm/rccl-tests commit: a6593375bc]
 2023-08-04 19:33:39 -04:00
Edgar Gabriel 1b21a6fdd8 search SLES install paths for MPI 
[ROCm/rccl-tests commit: 6048078be2]
 2023-07-24 12:02:44 -07:00
Wenkai Du 431f21ca11 Remove hardcoded number of GPUs limit for alltoallv (#41) 
[ROCm/rccl-tests commit: fcd0888d53]
 2023-06-18 18:07:29 -07:00
Wenkai Du a87b64d7c0 Fix merge error 
[ROCm/rccl-tests commit: 652a24d38d]
 2023-06-14 20:26:33 +00:00
Wenkai Du a2e8a44adb Merge remote-tracking branch 'nccl/master' into develop 
[ROCm/rccl-tests commit: bb0f15d407]
 2023-06-14 08:21:02 -07:00
Wenkai Du 7820b9f63f Merge remote-tracking branch 'origin/master' into develop 
[ROCm/rccl-tests commit: 469225bcaf]
 2023-06-14 08:01:50 -07:00
yangxingwu cba8bfd093 makefile: remove extra space 
[ROCm/rccl-tests commit: 52ea1b2148]
 2023-06-06 09:47:50 +00:00
Pedram Alizadeh aa5f75aa9c fixing the error message for mpirun when number of requested GPUs exceeds the limits (#37) 
[ROCm/rccl-tests commit: d16d1fb16b]
 2023-04-27 14:06:17 -04:00
Pedram Alizadeh b7054b500a Revert "fixing the error message for mpirun when number of requested GPUs exceeds the limits (#33)" (#36) 
This reverts commit 39f83d5bb7.

[ROCm/rccl-tests commit: e856fa720f]
 2023-04-25 13:44:43 -04:00
Pedram Alizadeh 39f83d5bb7 fixing the error message for mpirun when number of requested GPUs exceeds the limits (#33) 
[ROCm/rccl-tests commit: e146460810]
 2023-04-03 11:37:13 -04:00
alan.souza 4fd5ceeba8 fix handling of variable NVCC. Permit overriding the variable using environment variables 
[ROCm/rccl-tests commit: 7ccda3c97b]
 2023-03-25 16:56:16 -03:00
Felix Abecassis b3db782c3f Update README.md 
[ROCm/rccl-tests commit: 17d0a42d5a]
 2023-03-23 09:05:41 -07:00
Edgar Gabriel e86966907d revamp cmake MPI detection 
we honor user requested MPI installations using MPI_PATH first,
and check afterwards for MPICH and Open MPI in the default
Ubuntu and RHEL installation directories.


[ROCm/rccl-tests commit: bdf58b1656]
 2023-03-02 19:40:13 +00:00
Pedram Alizadeh cd256748c0 Adding -pthread flag for linking issues into CMakeLists.txt and src/Makefile (#31) 
[ROCm/rccl-tests commit: 255750b094]
 2023-03-02 11:05:25 -05:00
Pedram Alizadeh ed7707fa1b Adding -pthread flag for linking issues into src/Makefile (#30) 
* Adding -pthread flag for linking issues into src/Makefile

* Adding -pthread flag for linking issues into CMakeLists.txt

[ROCm/rccl-tests commit: 5275aa5715]
 2023-02-24 21:39:04 -05:00
Edgar Gabriel cc584c4825 auto-detect and enable MPI 
[ROCm/rccl-tests commit: 2b2f23f42d]
 2023-02-23 18:27:08 +00:00
Sylvain Jeaugey b70cac2b33 Update README.md 
Improve MPI example to avoid confusion of number of processes / total number of GPUs.

https://github.com/NVIDIA/nccl-tests/issues/54#issuecomment-1212023369

[ROCm/rccl-tests commit: 2cbb968101]
 2023-01-03 08:47:43 +01:00
David Addison 129a1b4b78 Add boot_id to the hostname hash due to collisions on Azure 
Fixes #60


[ROCm/rccl-tests commit: 0b4c4cb99f]
 2022-12-12 01:16:46 -08:00
Jithin Jose 5ba670d551 Use DJB2a hash algorithm in getHostHash() 
[ROCm/rccl-tests commit: 0aeba157db]
 2022-12-12 01:16:38 -08:00
Edgar Gabriel 84ee112d50 fix algorithm assigning values in testsuite 
avoid a division by zero which seems to only occur for op=prod and
datatype=half, since the maximum exponent is small (15) and can exceed
the number of ranks.


[ROCm/rccl-tests commit: e9f5be184c]
 2022-11-30 23:01:46 +00:00
David Addison 6313530fcc Call cudaFreeHost() on wrongPerGpu not cudaFree() 
[ROCm/rccl-tests commit: 24fcf64ed1]
 2022-11-22 11:18:37 -08:00
David Addison 04b5c40b1c Add fflush(stdout) before perf output 
[ROCm/rccl-tests commit: 3bd2bd292b]
 2022-11-22 11:16:47 -08:00
akolliasAMD 1db5cf9986 added std::max to avoid buffer overflow in printing (#25) 
[ROCm/rccl-tests commit: 9d3a53dfa3]
 2022-11-01 11:34:55 -06:00
Edgar Gabriel 3c2ff83a6a make cmake stage also pass in CI 
the subdir entry is not actually required for the compilation.


[ROCm/rccl-tests commit: 377b28e5fb]
 2022-10-31 22:07:15 +00:00
Edgar Gabriel 78706b1a95 add the rccl/lib directory to the link path 
[ROCm/rccl-tests commit: 9c9746739a]
 2022-10-31 19:01:22 +00:00
Edgar Gabriel 9b06fe6fd5 fix a messing endif statement 
error introduced with the web merger-resolution tool :-(


[ROCm/rccl-tests commit: 8a754f15ad]
 2022-10-25 16:31:57 +00:00
Edgar Gabriel d76b2db507 Merge branch 'develop' into topic/v2.13.4-sync 
[ROCm/rccl-tests commit: 4d7cd871c1]
 2022-10-21 17:12:45 -05:00
Wenkai Du 036145c75f Allow more precise measurements of single operation (#20) 
[ROCm/rccl-tests commit: 9a89c300b6]
 2022-10-21 22:07:41 +00:00
Edgar Gabriel fab39367ae make rccl-test compile again. 
all files compile now.
mpi tests also pass


[ROCm/rccl-tests commit: 641e93e99c]
 2022-10-21 22:07:33 +00:00
Edgar Gabriel 08f9435e5a Merge remote-tracking branch 'nccl-tests/master' into topic/v2.13.4-sync 
[ROCm/rccl-tests commit: 3ae371cce7]
 2022-10-14 16:02:54 -05:00
Wenkai Du d6264a0f11 Allow more precise measurements of single operation (#20) 
[ROCm/rccl-tests commit: d22281cb3f]
 2022-10-12 17:28:04 -07:00
Sylvain Jeaugey c0e3f4d443 Fix build on RHEL7 with GCC 4.8 
Add -std=c++11 to CXXFLAGS.
Fixes #116.


[ROCm/rccl-tests commit: 365b92a1ea]
 2022-10-12 01:24:14 -07:00
akolliasAMD 61fd36c006 removed hypercube from Makefile (#19) 
[ROCm/rccl-tests commit: 3fbd3280ce]
 2022-09-29 15:36:39 -06:00
Sylvain Jeaugey fdaa88710b Update NCCL tests 
[ROCm/rccl-tests commit: d313d20a26]
 2022-09-23 01:13:29 -07:00
David Addison 35ee4ec3eb Fix preprocessor version check for ncclGetLastError() 
ncclGetLastError() was added in NCCL 2.13.0


[ROCm/rccl-tests commit: 749573f2d6]
 2022-09-07 16:10:41 -07:00
David Addison a43863e1a7 Fix an issue with the last commit when data checking is disabled 
[ROCm/rccl-tests commit: afa4c56b6a]
 2022-09-07 11:23:49 -07:00
David Addison 59ed17798f Display N/A for error count in AlltoAll in-place test 
AlltoAll does not support in-place buffers


[ROCm/rccl-tests commit: a0a14911ee]
 2022-09-06 13:17:15 -07:00
John Bachan 70b6c0f5e5 Changed top-level Makefile behavior so that BUILDDIR is interpreted 
as relative to top-level directory. This done is by abspath'ing it before
passing it to subdirectory Makefile's.

The old behavior had two cases: with and without BUILDDIR being set by
the user. With BUILDDIR not set, the build dir would be named "build"
in the top-level directory. If BUILDDIR was set, then the build dir
would be placed at "src/${BUILDDIR}".

The new behavior is simpler, if BUILDDIR is not set then it defaults
to "build", and the directory holding the final build is always at just
"${BUILDDIR}" in the top level.


[ROCm/rccl-tests commit: bc5f7cfb0a]
 2022-08-23 10:08:49 -07:00
John Bachan b5d746b58e Resync with NCCL 2.13 
* Added "verifiable", a suite of kernels for generating and verifying reduction
  input and output arrays in a bit-precise way.
* Data corruption errors now reported in number of wrong elements instead of max
  deviation.
* Use ncclGetLastError.
* Don't run hypercube on non-powers of 2 ranks.
* Fix to hypercube data verification.
* Use "thread local" as the defaut CUDA capture mode.
* Replaced pthread_yield -> sched_yield()
* Bugfix to the cpu-side barrier/allreduce implementations.


[ROCm/rccl-tests commit: 51af5572bf]
 2022-08-22 17:51:06 -07:00
Wenkai Du afce63313a Fix typo from previous merge 
[ROCm/rccl-tests commit: 45ec598ac4]
 2022-08-12 14:42:17 +00:00
gilbertlee-amd d369c80f16 Enabling hipGraph codepath for future support (#18) 
[ROCm/rccl-tests commit: f6f3c44a7a]
 2022-08-09 16:45:27 -06:00
Wenkai Du 5cc7ff0931 Fix missing error checking for AllocateBuffs due to merge (#17) 
[ROCm/rccl-tests commit: 9025051bbb]
 2022-08-09 11:04:38 -07:00
Liam Wrubleski 10d8a890fb Add CMake files to build & package (#15) 
* Add CMake files to build & package

* Change build technique on CI

* Correct CI build command

[ROCm/rccl-tests commit: d704668bf7]
 2022-08-09 11:17:07 -06:00
Eiden Yoshida a8ad34e393 Allow gpu config override in CI (#14) 
[ROCm/rccl-tests commit: 2af4f6bc3a]
 2022-07-28 09:19:16 -06:00
akolliasAMD f6764628f4 updated alltoallV test to not have any zero values (#12) 
updated alltoallV test to not have any zero values between ranks

[ROCm/rccl-tests commit: 9925195afc]
 2022-07-21 10:28:53 -06:00
Edgar 69994bd15b update pytest before running CI 
There seems to be in an incompatibility between the python installation
used in the CI and pytest. Update pytest before running CI.


[ROCm/rccl-tests commit: 67544e2c34]
 2022-06-13 10:20:33 -04:00
Edgar dad6d819d0 implementation of multi-rank support in rccl-tests. 
[ROCm/rccl-tests commit: 0500f2f132]
 2022-06-10 14:54:10 -04:00
Edgar b199d173a2 create branch up-to-date with rccl-test 
[ROCm/rccl-tests commit: 5cd2374edb]
 2022-06-10 12:41:56 -04:00
amdkila 652d1963af Check for error code in install script (#2) 
[ROCm/rccl-tests commit: 3d6f70659a]
 2022-06-10 12:37:53 -04:00